Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/tideshift-labs/fathom/agents-mdgit clone --depth 1 https://github.com/Tideshift-Labs/FathomWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00683 | $0.00683 |
| Opus 5 | $0.00342 | $0.00342 |
| Sonnet 5 | $0.00137 | $0.00137 |
| Haiku 4.5 | $0.00068 | $0.00068 |
Grade A, and why
Fathom AGENTS.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 46 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent Instructions: Fathom (Rider Plugin)
Style
- Never use emdashes (
—). Rephrase instead.
Key Documentation
- README.md: Architecture, HTTP API, build commands, development workflow.
- docs/LEARNINGS.md: Dead-end approaches and hard-won lessons. Read the "DEAD END" sections before attempting alternative inspection APIs. Multiple approaches (RunLocalInspections for C++, ForceReHighlight for non-open files, SolutionAnalysisManager as a component) were fully explored and proven not to work.
- docs/ue-companion-plugin.md: Blueprint audit architecture, explaining how the Rider plugin and UE companion plugin couple via filesystem conventions.
Reflection Gotchas
This plugin accesses closed-source JetBrains APIs via reflection. Key pitfalls:
GetComponent<T>()is an extension method, not an instance method onISolution. You must search interfaces, base types, then scan assemblies for static extension methods.ResolveComponent(Type)helper exists for component resolution when generic type parameters don't work at runtime.- Type names may differ from guesses. Use diagnostic dumps (stderr or HTTP response with
&debug=true) when debugging reflection failures. - Blueprint APIs (
UE4AssetsCache,UE4SearchUtil) are accessed via reflection onJetBrains.ReSharper.Feature.Services.Cpp.dllwith no compile-time reference.
UE Project Detection
Two approaches are used, depending on the need:
- Lightweight: Check for
.uprojectfiles in the solution directory. Fast, no reflection needed. - Full info:
GetUEProjectInfo()viaICppUE4SolutionDetectorreflection. Returns engine path, project path, target platform. Engine path comes fromUnrealContext.Pathproperty.
Commandlet Detection
When the Rider plugin spawns UnrealEditor-Cmd.exe -run=BlueprintAudit, it looks for "unknown commandlet" patterns in the process output to detect whether the companion UE plugin is installed. If the commandlet isn't registered (plugin not installed), the output contains this pattern and the refresh fails gracefully.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 3d ago First seen · 46 lines · 683 tokens per session scan A 036df4907fd7
Fathom AGENTS.md is an instructions file published in the GitHub repository Tideshift-Labs/Fathom (14 stars, last pushed 1mo ago), licensed MIT. It adds 683 tokens to every session, about $0.0034 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
Unreal_mcp AGENTS.md
Instructions for ChiR24/Unreal_mcp, covering project knowledge base, structure, where to look, code map and conventions.
ue-mcp CLAUDE.md
Instructions for db-lyon/ue-mcp, covering claude.md, repo at a glance, development workflow, git and commits and versioning.
Unreal_mcp copilot-instructions.md
Instructions for ChiR24/Unreal_mcp, covering copilot instructions for unreal mcp, architecture, critical constraints, ue 5.7 safety and build and test.
unreal-mcp CLAUDE.md
Instructions for GenOrca/unreal-mcp, covering claude.md, what this is, architecture (the non-obvious parts), commands and all run from mcp-server/.
loomle AGENTS.md
Instructions for loomle/loomle, covering loomle working rules, goal, public protocol naming, mutation dry run contract and workflow.
mcp-unreal CLAUDE.md
Instructions for remiphilippe/mcp-unreal, covering claude.md — mcp-unreal, project overview, key facts, repository structure and code standards.